1 / 34

Nick Gotelli University of Vermont USA

Detecting Temporal Trends In Species Assemblages With Randomization Procedures And Hierarchical Models . Nick Gotelli University of Vermont USA. Collaborators!. Robert Dorazio University of Florida USA. Gary Grossman University of Georgia USA. Aaron Ellison Harvard Forest USA.

mireya
Download Presentation

Nick Gotelli University of Vermont USA

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Detecting Temporal Trends In Species Assemblages With Randomization Procedures And Hierarchical Models Nick Gotelli University of Vermont USA

  2. Collaborators! Robert Dorazio University of Florida USA Gary Grossman University of Georgia USA Aaron Ellison Harvard Forest USA

  3. Causes of Temporal Change in Communities

  4. Pathways of Temporal Change Abiotic Change Changes in abundance Changes in abundance of competitors, predators, prey

  5. Conspicuous Drivers of Temporal Change • Keystone Species • Foundation Species • Ecosystem Engineers • Invasive Species

  6. Subtle Drivers of Temporal Change • Habitat alteration, succession • Long-term climate change • Hunting, overexploitation • “Shifting Baseline”

  7. But not all apparent patterns of temporal change reflect “true” changes in population or community structure!

  8. Most indices of species diversity and population size are sensitive to “sampling” effects

  9. How can we account for sampling effects when assessing temporal changes in populations and communities?

  10. Data Structure i = 1 to S species j = 1 to T consecutive temporal samples yij = count of individuals of species i recorded in sample j

  11. Freshwater fishes in a central U.S. stream i= 1 to 55 species j = 1 to 15 ~ annual samples (1963 – 1974) N = 14,142 individuals sampled by seining Grossman, G. D., Moyle, P. B., and J. R. Whitaker, Jr. 1982. Stochasticity in structural and functional characteristics of an Indiana stream fish assemblage: a test of community theory. Am. Nat. 120:423-454.

  12. Insects in a central U.S. grassland (KBS) i= 1 to 9 species common species (Chrysopidae, Lampyridae ) j = 1 to 14 annual samples (1989 – 2002) N = 5614 individuals sampled by sticky traps Isaacs, R., J. Tuell, A. Fiedler, M. Gardiner, and D. Landis. 2009. Maximizing arthropod-mediated ecosystem services in agricultural landscapes: The role of native plants. Frontiers in Ecology and the Environment 7: 196-203.

  13. Null model test for temporal trends in community structure • Metric to summarize pattern of temporal change (TC) • Specify distribution of TC under sampling H0

  14. Abundance Trends For A Single Species

  15. Abundance Trends For A Single Species

  16. Abundance Trends For A Single Species

  17. Community Trends in Abundance Stationary Non-Stationary Null hypothesis for measurement of temporal trends at community level

  18. Metric to summarize pattern of temporal change TC is the sample variance of trend line slopes for all species in the assemblage

  19. Community Trends in Abundance Stationary Non-Stationary

  20. Specify distribution of TC under sampling H0 • Assign each of individuals N to different time periods based on tj, the proportion of the total collection made at time j (good and bad sampling intervals) • Assign each of the N individuals to a different species based on pi, the proportion of the total collection represented by species i (common and rare species)

  21. Assumptions of Null Model • Multinomial sampling, conditional on total abundance (N) • Species differ in commonness and rarity • Time periods differ in suitability for detection • No species interactions

  22. Incorporating Undetected Species • Observed S is a biased under-estimator of total S • Undetected species should be included in the null distribution • Estimate the number of missing species using non-parametric Chao2 estimator (Chao 1984)

  23. Non-parametric Estimator for Undetected Species T = number of censuses Q1 = number of “singletons” (species detected in exactly 1 census) Q2= number of “doubletons” (species detected in exact;u 2 censuses) Chao, A. 1984 Non-parametric estimation of the number of classes in a population. ScandinavianJournal of Statistics 11: 265-270.

  24. Estimating Relative Abundance

  25. Estimating Relative Abundance Undetected Species

  26. Estimating Relative Abundance Assumption: Relative frequency of undetected species = 0.5 x relative frequency of rarest observed species Undetected Species

  27. Temporal Trends of Stream Fishes Total Abundance (1963-1974)

  28. Temporal Trends of Stream Fishes Individual Species (1963-1974) Null Distribution

  29. Temporal Trends of Grassland Insects Total Abundance (1989-2002)

  30. Temporal Trends of Grassland InsectsIndividual Species (1989-2002) Null Distribution

  31. Estimating Temporal Trends For Individual Species • Assumes model of exponential growth • Poisson distribution for population size • Detection probabilities differ among species, but are constant across sampling dates • Growth rates for individual species estimated from common distribution • Model cannot be fit for species that are very rare (< 10 occurrences)

  32. Estimated Growth Rates of Stream Fishes

  33. Estimated Growth Rates of Grassland Insects

  34. Summary • Temporal changes in community structure generated by abiotic forces and species interactions • Multinomial sampling model as a null hypothesis for temporal trends • Heterogeneous patterns forstream fishes and grassland insects • Hierarchical model to estimatetrends for individual species

More Related